Tikal2k - here is a link to the current series of FIB (actually dual beam) systems used to do this specific work. (although this is probably on an older series of FIB system - universities usually dont have the cash to buy new ones)

http://www.smt.zeiss.com/C1256E4600305472/Contents-Frame/40C720DCA03B0C6CC1256E5400416DEE

FEI and DCG Systems are other manufacturers of commercial FIB systems

The fact that it is planar leads more to its authenticity anyway. The only way to have it mill "correctly" into a curved or cylindrical surface is to adjust the final lens voltage proportional to the curvature of the surface. A LEO (Zeiss) FIB cant easily do this.

The two etched images are due to the smaller one being "washed out" because the beam spot size is too large. This was their "practice" bitmap mill. Notice how you can't make out the wings in the logo? Kinda like drawing a very small picture with an dull pencil - draw the same picture in a bigger scale with the same pencil and it will look much better.

The horizontal raster striations are due to the beam not being perfectly astigmated + raster artifacts.

Scientists at Los Alamos (in the late 90's)already patented a way to archive data using a focused ion beam and reading it with an AFM.

Prewett and Mair have a similar bitmap driven FIB image on the cover their 1991 book.

Here are some others (along with some FIB science) by just doing a quick google search using "fib logos" - http://www.s3.infm.it/fib_index.html